This paper evaluates the biological regeneration of ferric Fe3+ solution during desulphurisation of gaseous streams. Hydrogen sulphide (H2S) is absorbed into aqueous ferric sulphate solution and oxidised to elemental sulphur, while ferric ions Fe3+ are reduced to ferrous ions Fe2+. During the industrial regeneration of Fe3+, nutrients and trace minerals usually provided in a laboratory setup are not present and this depletion of nutrients may have a negative impact on the bacteria responsible for ferrous iron oxidation and may probably affect the oxidation rate. In this study, the effect of nutrients and trace minerals on ferrous iron oxidation have been investigated and the results showed that the presence of nutrients and trace minerals affects the efficiency of bacterial Fe2+oxidation. The scanning electron microscopy analysis of the geotextile support material was also conducted and the results showed that the iron precipitate deposits appear to play a direct role on the bacterial biofilm formation.

INTRODUCTION

Hydrogen sulphide is one of the major air pollutants and is highly toxic, malodorous and corrosive (Kim et al. 2002). Although many commercial processes are available for the removal of H2S from gaseous streams, the desulphurisation process using aqueous ferric sulphate solution as an absorption medium is a cheaper and cleaner alternative as only water and sulphur (targeted product) are generated. H2S is absorbed into aqueous ferric sulphate solution and oxidised to elemental sulphur, while Fe3+ is reduced to Fe2+. Elemental sulphur can be removed from the solution in a separate reactor; the reactant Fe3+ is regenerated from the Fe2+ solution by biological oxidation in an aerated bioreactor using bacteria such as Thiobacillus ferrooxidans and Leptospirillum ferrooxidans. The biological oxidation of ferrous iron to ferric iron is the rate-limiting step (Kim et al. 2002; Hallberg et al. 2011).

Micro-organisms such as Thiobacillus ferrooxidans are chemolithotrophic as their energy for growth and maintenance is derived from the oxidation of ferrous iron to ferric iron. They are aerobic requiring oxygen as an electron acceptor and are autotrophic using carbon dioxide (CO2) as a cellular carbon source. The ferrous iron-oxidising bacteria are acidophilic and grow within the pH range of 1.5–6.0; however the optimum growth occurs at pH 2.0 and 2.5 (Leduc & Ferroni 1994; Amouric et al. 2010).

During desulphurisation, when H2S is absorbed into aqueous ferric sulphate solution and oxidised to elemental sulphur, and Fe3+ is generated, acidity is formed according to the following equation: 
formula
1
This acidity generating reaction may lower the pH to an extent where the performance of the biological process of regenerating Fe3+ from the Fe2+ is affected. When Fe2+is oxidised biologically to Fe3+, the pH value initially increases, since oxidation of ferrous iron is accompanied by removal of acid (Jensen & Webb 1995). This reaction occurs as follows: 
formula
2
The reaction depletes O2, H+ and Fe2+, while producing Fe3+, H2O and energy needed to fix CO2 during cellular growth (Pagella et al. 1996; Schippers et al. 2010). This reaction may allow for sufficient buffering of the solution pH, but may not be adequate to compensate for the acidity produced during reaction 1. During the regeneration of Fe3+, nutrients and trace minerals usually provided with synthetic feed will not be present and this depletion of nutrients may have a negative impact on the bacteria responsible for ferrous iron oxidation and affect the reaction rate. The effect of nutrients and trace minerals on ferrous iron oxidation is, therefore, investigated in this study.

EXPERIMENTAL

Feed water

The chemical composition of the synthetic feed water is presented in Table 1. The pH of the feed water was adjusted to pH 2. The start Fe2+ concentration was approximately 3 g/L of Fe2+ during the different experimental periods.

Table 1

The chemical composition of the Fe2+ synthetic feed water (Maree & Van Tonder 1999)

Chemical Amount (g/L) 
FeSO4.7H215 
KH2PO4 0.18 
MgSO4.7H20.23 
NH4SO4 0.23 
H2SO4 2.04 
Chemical Amount (g/L) 
FeSO4.7H215 
KH2PO4 0.18 
MgSO4.7H20.23 
NH4SO4 0.23 
H2SO4 2.04 

Reactor system

Five L fluidised bed reactors containing geotextile as the support material for the bacteria to adhere to were used to investigate the iron oxidation process as shown in Figure 1. Seven fresh geotextile strips (3.5 × 170 cm) were added to the reactors. The reactors were operated in sequential batch mode throughout the experimental periods. The feed water in the reactor was continuously recycled from the top of the reactor to the bottom of the reactor by using a Watson Marlow pump 323 set at 20 mL/min. Since the Fe2+ to Fe3+ oxidation is an aerobic process, air was provided for the reactors through a venture system, which was delivered with the recycling water stream. The airflow to the reactor was constant at 9 L/min. The temperature of the reactor was maintained at 30 °C using an aquarium heater. As very high levels of H2S were used in early experiments, prior to using the regenerated Fe2+ solution in the biological reactor, N2 gas was bubbled through the solution.

Figure 1

Experimental setup.

Figure 1

Experimental setup.

Micro-organism and culture medium

The bacterial strain used in this study (T. ferrooxidans) was obtained from the Navigation coal mine water, South Africa. T. ferrooxidans culture was maintained on synthetic feed water as shown in Table 1 and inoculated with 5% (vol/vol) acid mine drainage from the Navigation coal mine. The bacterial cultures were activated and maintained at 30 °C using a bacterial incubator shaker at 150 r/min.

Scanning electron microscopy

Geotextile samples were fixed in 10 mL of 2.5% gluteraldehyde and 10 mL of 0.075 M NaPO4 buffer for 10 min at room temperature in triplicate. The samples were then washed with the NaPO4 buffer for 15 min in triplicate. The samples were dehydrated in a series of 50, 70, and 90% ethanol for 15 min each followed by 15 min in 100% ethanol in triplicate. The samples were then critical point dried, coated with gold and then viewed under a JEOL-840 scanning electron microscope at an acceleration voltage of 5–15 kV.

Analyses

Measurements of pH, and DO were conducted according to Standard Methods (1985). The Fe2+ measurement method was a titration method according to Jeffery et al. (1989).

RESULTS AND DISCUSSION

The effect of nutrients on the oxidation of Fe2+ to Fe3+

Bacteria such as Thiobacillus ferrooxidans that can oxidise ferrous iron to ferric iron are autotrophic, using carbon dioxide as the cellular source of carbon (Jensen & Webb 1995). N and P are also required as nutrients for cellular growth and synthesis as well as trace minerals of K, Mg, Na, Ca and Co (Jensen & Webb 1995). Most of these nutrients and trace minerals are present in the Fe(II) synthetic feed (Table 1). During the desulphurisation process, when H2S is absorbed into aqueous ferric sulphate solution and oxidised to elemental sulphur, and Fe3+ is generated, these nutrients and trace minerals will not be present and this may have a negative impact on the bacterial ferrous iron oxidation reaction rate. The effect of nutrients was observed in two reactors: reactor A and B. Reactor A was fed synthetic feed with the nutrients according to Table 1 at a pH of 2. Reactor B was fed synthetic feed containing only the FeSO4 and H2SO4, without the nutrients. Prior to the experiment it was determined that the oxidation rates of both reactors A and B were very similar and these reactors could therefore be used for comparison studies.

The effect of nutrients was examined over six iterations (Figures 2(a)2(f)). An iteration refers to the experimental period where new feed is added to the reactor and the Fe2+oxidation to Fe3+ is monitored. Although both reactors A and B showed similar oxidation rates in the first iteration (Figure 2(a)), from the second iteration reactor B began to oxidise the Fe2+at a slower rate (Figure 2(b)). The rate of Fe2+oxidation gradually declined with each successive iteration when compared to reactor A (Figures 2(c)2(f)). The average Fe2+ to Fe3+ oxidation for all six iterations is shown in Figure 3 with the comparison values. The Fe2+ oxidation rate of reactor A remained fairly constant for iterations 1–6, while reactor B showed a gradual decline in Fe2+ oxidation from 15.01 to 8.23 g/L d (Figure 3). This decline in performance indicates that the presence of nutrients and trace minerals affects the efficiency of bacterial Fe2+ oxidation. It has been shown that the iron precipitates during the bacterial oxidation of ferrous iron are mainly potassium jarosite and that potassium ion plays a major role in precipitation (Grishin & Tuovinen 1988). The potassium ion concentrations can possibly be used to manipulate the amount of precipitates in the reactor (Kinnunen & Puhakka 2004). Since the iron precipitates or deposits appear to play a direct role in the process of bacterial biofilm formation (Karamanev 1991; Nemati & Webb 1999), this highlights the need for the inclusion of trace minerals and nutrients in the system.

Figure 2

The oxidation of Fe2+ to Fe3+ in reactor A (with nutrients) and reactor B (without nutrients) depicting iterations: (a) 1, (b) 2, (c) 3, (d) 4, (e) 5 and (f) 6.

Figure 2

The oxidation of Fe2+ to Fe3+ in reactor A (with nutrients) and reactor B (without nutrients) depicting iterations: (a) 1, (b) 2, (c) 3, (d) 4, (e) 5 and (f) 6.

Figure 3

The average Fe2+ oxidation in g/L d from reactor A (with nutrients) and reactor B (without nutrients) with standard deviation.

Figure 3

The average Fe2+ oxidation in g/L d from reactor A (with nutrients) and reactor B (without nutrients) with standard deviation.

Eight iterations were also conducted on an industrial ferrous iron solution generated after absorption with H2S in the biological reactor to regenerate Fe(III). The pH of the regenerated solution was not adjusted prior to adding to the biological reactor, but remained within the region of pH 1.91–2.37. On average the reactor oxidised 5.09 g/L d of Fe(II) in the eight iterations observed. This is a much lower rate than that obtained when reactor A was operated with synthetic feed, which was 13.67 g/L d over the six iterations observed in Figure 3. The regenerated Fe(II) solution was measured for the presence of sulphides, which may be inhibitory to the bacteria responsible for ferrous iron oxidation, however the concentration of sulphides in the regenerated Fe(II) solution was <148 mg/L. The regenerated Fe(II) solution did not contain nutrients and this could also explain the slower removal rate of ferrous iron compared to the synthetic feed. The importance of including trace minerals and nutrients is therefore reiterated in this experiment.

The scanning electron microscopy of the support material

The use of support material for the micro-organisms is especially important under continuous flow conditions in order to prevent cell wash-out. Iron-oxidising bacteria have a propensity to grow on surfaces and this property has been exploited in bioreactors with various immobilisation matrices ensuring a stable biomass is retained (Ebrahimi et al. 2005; Nurmi et al. 2009). In a study investigating biological oxidation of iron in acid mine water, it was determined that geotextile was the best support media for optimal iron oxidation, resulting in accelerated bacterial absorption and biofilm formation. The porosity of the geotextile allows the air to penetrate into the fibres easily, providing a large surface area for the bacteria to adhere to (Nengovhela et al. 2004). Thiobacillus ferroxidans is the bacteria commonly responsible for Fe2+ oxidation (Nemati et al. 1998). Biofilm is defined as an assemblage of microbial cells associated with a surface and enclosed in a matrix of primarily polysaccharide material (Donlan 2002). The scanning electron microscopy revealed that the biofilm on the geotextile carrier material was covered with jarosite precipitates (Figures 4(a)4(f)). Jarosite is a porous inorganic substance with the general chemical formula of AFe3(SO4)2(OH)6 where A can be K+, Na+ NH4+, or H3O+ (Nicolov et al. 2002). Individual bacterial cells could not be visualised on the biofilm surface as the immobilised bacteria were embedded in the aggregates of ferric iron precipitates. This is in agreement with Nemati & Webb (1999), where layers of bacteria on the biofilm support particles were embedded in an aggregate of ferric iron precipitates and some of them were completely covered by these compounds. The jarosite deposits appear to play a direct role in the process of biofilm formation (Karamanev 1991; Nemati & Webb 1999) and microbial attachment onto carriers is influenced by the increasing quantities of precipitated jarosite (Pogliani & Donati 2000; Kinnunen & Puhakka 2004; van der Meer et al. 2007). In addition to accelerating the adsorption of cells, ferric iron precipitates may also play an important role in the mechanical strength of the biofilm (Nemati & Webb 1999). In the classic theory of biofilm structure bacteria form extracellular polymeric substances, which bind the cells to the attachment surface and each other. It has been suggested that T. ferrooxidans cells themselves are not part of the mechanical structure of the biofilm, but are absorbed as a monolayer to the framework of the biofilm made of jarosite (Karamanev 1991). The biofilm on the geotextile appeared as a thick conglomerates of ferric iron precipitate, which may play a role in the strength of the biofilm and its ability to withstand the shear forces of the air flow provided (Figures 4(d)4(f)). The jarosite has a porous structure (Nicolov et al. 2002), promoting microbial attachment and forms the framework of the biofilm which is independent of the microbial growth rate (Karamanev 1991).

Figure 4

The biofilm on the geotextile fibres serving as support material showing the crystal structure of the ferric iron precipitate.

Figure 4

The biofilm on the geotextile fibres serving as support material showing the crystal structure of the ferric iron precipitate.

CONCLUSIONS

The aim of this study was to examine the parameters affecting the regeneration of reactant Fe3+ from Fe2+ solution by biological oxidation in an aerated bioreactor. The effect of nutrients and pH on biological Fe(II) oxidation to Fe(III) was investigated and the rate of Fe(II) oxidation was observed when using regenerated Fe(II) after H2S absorption. The Fe(II) oxidation rate was on average 13.67 g/L d when operated with synthetic feed containing nutrients and 10.93 g/L d when operated with synthetic feed without nutrients, however it is likely that this value would continue to decline with further operation as the Fe(II) oxidation rate decreased with each successive iteration as remaining nutrients were washed out of the system. The laboratory results show that it is essential that nutrients be supplemented when iron is regenerated for the purpose of H2S removal to ensure optimal performance of the bacteria responsible for this reaction. When an industrial sample of regenerated Fe(II) was oxidised to Fe(III) after H2S absorption, the reactor oxidised on average 5.09 g/L d of Fe(II). This oxidation rate was much lower than that obtained when the reactor was operated with synthetic feed, which may have been due to the lack of nutrients in the regenerated Fe(II) solution, which reinforced the significance of including trace minerals and nutrients. The biofilm structure on the geotextile carrier material and scanning electron microscopy revealed that the biofilm on the geotextile carrier material was covered with jarosite precipitates, which play a vital role in bacterial biofilm development.

REFERENCES

REFERENCES
Donlan
R. M.
2002
Biofilms: microbial life on surfaces
.
Emerging Infectious Diseases
8
(
9
),
1
19
.
Ebrahimi
S.
Fernández Morales
F. J.
Kleerebezem
R.
Heijnen
J. J.
van Loosdrecht
M. C. M.
2005
High-rate acidophilic ferrous iron oxidation in a biofilm airlift reactor and the role of the carrier material
.
Biotechnology and Bioengineering
90
(
4
),
462
472
.
Grishin
S. I.
Tuovinen
O. H.
1988
Fast kinetics of Fe2+ oxidation in packed-bed reactors
.
Applied and Environmental
54
(
12
),
3092
3100
.
Hallberg
K. B.
Grail
B. M.
du Plessis
C. A.
Johnson
D. B.
2011
Reductive dissolution of ferric iron minerals: a new approach for bio-processing nickel laterites
.
Mineral Engineering
24
,
620
624
.
Jeffery
G. H.
Bassett
J.
Mendham
J.
Denney
R. C.
1989
Vogel's Textbook Quantitative Chemical Analysis
,
5th edn
.
Publisher: John Wiley & Sons, Inc.
,
New York
.
Jensen
A. B.
Webb
C.
1995
Ferrous sulphate oxidation using Thiobacillus ferrooxidans: a review
.
Process Biochemistry
30
(
3
),
225
236
.
Karamanev
D. G.
1991
Model of the biofilm structure of Thiobacillus ferrooxidans
.
Journal of Biotechnology
20
,
51
64
.
Leduc
L. G.
Ferroni
G. D.
1994
The chemolithotrophic bacterium Thiobacillus ferrooxidans
.
FEMS Microbiology Reviews
14
(
2
),
103
120
.
Maree
J. P.
Van Tonder
G. J.
1999
New Development in Neutralisation of Acid Mine Water. Chemical Technology
,
Issue Nov–Dec, Crown Publications, Johannesburg
.
Nengovhela
N. R.
Strydom
C. A.
Maree
J. P.
Greben
H. A.
2004
Chemical and biological oxidation of iron in acid mine water
.
Mine Water and the Environment
23
,
76
80
.
Nicolov
L.
Karamanev
D.
Mamatarkova
V.
Mehochev
D.
Dimitrov
D.
2002
Properties of the biofilm of Thiobacillus ferrooxidans formed in a rotating biological contractor
.
Biochemical Engineering Journal
12
,
43
48
.
Nurmi
P.
Ozkaya
B.
Kaksonen
A. H.
Tuovinen
O. H.
Riekkola-Vanhanen
M.-L.
Puhakka
J. A.
2009
Process for biological oxidation and control of dissolved iron in bioleach liquors
.
Process Biochemistry
44
,
1315
1322
.
Pagella
C.
Perego
P.
Zilli
M.
1996
Biotechnological H2S gas treatment with Thiobacillus ferrooxidans
.
Chemical Engineering Technology
19
,
79
88
.
Standard Methods for the Examination of Water and Wastewater
.
1985
16th edn
.
American Public Health Association
,
Washington, DC, USA
.
Van der Meer
T.
Kinnunen
P. H.-M.
Kaksonen
A. H.
Puhakka
J. A.
2007
Effect of fluidized-bed carrier material on biological ferric sulphate generation
.
Minerals Engineering
20
,
782
792
.